ML19347B864
| ML19347B864 | |
| Person / Time | |
|---|---|
| Issue date: | 09/04/1980 |
| From: | Catton I Advisory Committee on Reactor Safeguards |
| To: | Bates A Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1282, NUDOCS 8010160114 | |
| Download: ML19347B864 (4) | |
Text
-
UN15EVWunof ADVIS% COMYmEE ON LOS ANGELES. SCHOOL OF ENGINEERING AN A PPLIED SCIENCE REACTOR SAf[GUAOS, U1h R.C.
" '~ M N 4
SEP 15880
[,N 3 b JJ,12 4 2 g3 /~ M W
4 Sept.1980 I
TO:
A. Batss FROM:
- 1. Catton
$d
SUBJECT:
ORNL Blow Down Heat Transfer Program ORNL bas redirected their program significantly since we last heard from The redirection occurred over a year ago and is now very responsive them.
to NRR needs.
It does appear, however, that THTF will have yielded most of what it is capable of when the present test matrix is completed.
There is some potential for future use of THTF in other areas.
The ORNL staff will have to become a great deal more aware of what is going on in the field than they are now to determine how their facility can help fill gaps in our knowledge.
I recommended they look into areas such as DNBR.
The following paragraphs contain a summary of my views and some detailed discussion of several aspects of the ORNL program.
SUMMARY
1.
The present THTF facility is much better than the previous facility.
It is well instrumented and appears to have the capability of being very productive over the next twelve months.
Some atypicalities such as the high heater pin thermal conductivity, the close spacing of grid spacers, and the heavy shroud wall detract from its potential and will require attention when use is made of the data.
2.
The BDHT program has undergone extreme redirection during the past year The large break LOCA no longer receives any attention.
The pre-or so.
sent program matrix will satisfy 80% of the needs expressed by NRR in their August 1979 letter.
It is doubtful if the remaining 20% can be addressed with THTF.
3.
Recent experimental results for forced convection to superheated steam, bundle recovery and film boiling heat transfer appear to have yielded valuable data.
Caution must be exercised to not be mislead by some of the preliminary conclusions reached by ORNL in interpreting their results.
Further efforts in reducing the data' are 'necessary if its full value is to be realized.
4.
Studies of the uncertainties in code predictions at ORNL should be dis-couraged.
They should obtain such information from Dr. Fabic through his code verification efforts.
5.
Use of the THTF facility by WRSR in satisfying NRR needs appears to be reaching a point of diminishing returns.
The facility may be able to contribute in other areas such as studies of DNBR and post DNB at ele-vated pressures and possibly ATMS. The ability of THTF to operate at very high pressures (P > 2000 psia) and high temperature needs to be better established.
QSOFFICE00Py at 4)pmawwm
FACILITY 4
A significant upgrading of THTF has taken place since we last heard about It is an entirely n'ew bundle with 64 rods of a new design.
the facility.
i The new desigr. appears to be vastly superior to the previous one although some negative asp.ects still exist. A Boron Nitride filler is used with an inner The filler is inside the incenel heater cylin-cylinder incenel heater element.
der ac d in the annulus between the heater and the stainless steel sheath.
Thermocouples are placed azimuthly around the inner surface of the stainless steel sheath and on the heater centerline.
The high conductivity of the Boron Nitride filler is both a help and a hindrance. Surface heat flux variation was found to be less than 1%. The The intimate can-pins are capable of up to 20 kw/ft with a cosine profile.
tact of the boron nitride with the stainless steel clad will have an impact on the quench front velocity relative to that of a fuel pin. The high thermal conductivity interior and intimate contact will allow more energy to be con-ducted to the quench front and in effect the pin will act as if it has a very thick cladding. This will cause a significant decrease in the quench front velocity.
In-bundle densitometers have been designed and are to be installed in the near future.
If they work as well as anticipated, the data obtained could be very valuable. Hopefully they will be installed soon.
Consideration should One will finally be given to their use in other facilities such as Semiscale.
be able to obtain most of the fluid conditions without resorting to a computer code.
Code verification will be less circuitous and correlations of the data will be on a much sounder basis.
The bundle has 960 thermocouples located on the inner surface of the heater pin sheaths, Thermocouples are located immediately above and below the grid spacers to establish grid effects. The azimuthal location of the thermocouples is lost when the pin stress relieves itself on heating.
It is not clear why the heater rods were not heat treated before installation.
Oversights such as this can be costly in terms of lost information. THTF is a very well instru-mented system.
The ORNL staff claim that they can run six tests per day and that any given test takes about two weeks to set up. This is a remarkable and probably over-ambitious claim. They have not performed this well to date. In any event, THTF appears to be a useful piece of equipment.
PROGRAM REDIRECTION The ORNL Blow Opwn Heat Transfer (BDHT) program has undergone major revision.
The program now appears to be very responsikt to the NRR request for confirma-tory research on LWR heat transfer (Denton to L,evine, Aug. 17, 1979). The Three of the letter emphasizes a number of aspects of the small break LOCA.
four priority one items in the letter are in part covered by the capabilities The present test of the facility and NRR expects THTF to answer these needs.
matrix will supply 80% of the information requested by NRR under its present UNIVERSIT Y OF C ALIFORNI AHLetterhead for interdepartmental use)
-a..
awumm.mmmm.
The remaining 20% appears to be beyond the capabilities of the facility.
I program.
RECENT EXPERIMENTAL RESULTS Forced Convection to Superheated Steam.
Steam cooling experMents have bero conducted and comparisons made with vendor models.
The comparison with vendor models showed 13-28% standard error. All overpredicted at high Reynolds None of the models correctly predict the effect of variations of the number.
ratio of rod to steam temperature ratio. The latter could be a result of the radiation correction made during the data reduction process.
The handling of the radiative contribution to the heat transfer during the data reduction process is primitive. Personnel seemed unaware of past research on the radiation properties of steam.
It was recommended that they do a little research on radiation in rod bundles.
In particular, they should contact Tien, at UCB, who has developed a model. With the primitive modeling done by ORNL, it is not clear their conclusions about the effect of a cold shroud are valid.
Further stainless steel will oxidize during testing. This will cause the emissivity to change with time. Without measurements during a given test series, it is difficult to argue that there is only a 20% uncertainty in the value of surface emissivity.
The ORNL staff indicated a desire to do more testing in this area with more emphasis on the radiation aspects of the problem.
It was my view that single phase convective cooling was well enough understood and that until a better model of the radiation contribution was used in the data reduction, future plans are premature. Further, if the radiation contribution deserves more attention, then companion experiments of a simpler nature are needed to develop the physics with THTF being used in a confirmatory fashion only.
Bundle Recovery Tests. The bundle recovery experiments will yield data that has relevance to the Small Break LOCA. Heat transfer and quench front behavior were measured. They noted the effects of pressure on the liquid level swell. When the in-bundle densitometers are installed much more meaningful data will result. A repeat of one or two tests with installed densitometers would be worthwhile.
It is worthwhile noting however that the maximum swell was 68% at 388 psia and only 6.5% at 1000 psia. Comparison with analytical level swell studies by Duffey and Sun (EPRI) would allow closure of this area of experimentation.
Results i
The reported results 6n quench front behavior are incomplete.
are shown that imply a pressure effect on quench front velocity when it may The ORNL measured quench front velocities were much j
be a temperature effect.
lower than those predicted by vendor models, and the conclusion that the vendor models may be incorrect is reached.
The low quench front velocities are not surprising when one considers the design of the heater rod and the high ther-The quench front velocity was typically mal conductivity of the Boron Nitride.
1/2 the flooding velocity. This is not dissimilar to UCLA results using solid steel pins.
It is premature to cast suspicion on the vendor models. A detailed analysis of the THTF heater pin quench front behavior using vendor models would be required before such conclusions can be reached.
Further, fuel pin gap UNIVERSITY OF CALIFORNI A-Hktterhead for interdepartmental use)
concoctance and low fuel thermal conductivity allow certain simplifications
- that probably can't be made in an analysis of the THTF heater pins.
1 Film Boiling Heat Transfer.
Film boiling heat transfer was measured for 4
a range of conditions and compared to correlations presently in use in the indus try. The data looks good and the comparison with some of the correla-tions show they are surprisingly good. Some further evaluation is needed, however.
In the steam cooling studies a radiation correction was made in obtaining a convective heat transfer coefficient.
In this portion of their work, it is not clear that they did so. The radiation contribution may be quite impor-tant and even larger than during steam cooling because the sink temperature is lower.
It was found that the Dougall-Rosenhow correlation overpredicted the film boiling heat transfer coefficient by a factor of two whereas the Groenveld correlations yielded very good comparison. First, if one assumes that radiation i
is not very important then the results may be reasonable if one considers that the Dougall-Rosenhow correlation was based on an analysis and adjusted to fit data from Freon tests inside a tube at relatively low flow rates, Rather than conclude a correlation is not any good, one should note that it should not be applied outside its range of applicability. The Groenveld correlations were based on steam-water data from tests having parameter ranges bracketing the regimes of interest.
Reaching conclusions about correlations in use in reactor safety analysis j
is serious business.
Before doing so, an exhaustive assessment of ones data needs to be made. As mentioned earlier, ORNL seems to have ignored radiation.
A very rough calculation indicates that the radiation contribution could be 41% of the total heat transfer. This makes the correlations even less con-servative unless the radiation contribution is built into the correlations in use.
One needs to review how they were developed and whether or not radia-tion corrections were made by Dougall & Rosenhow or by Groenveld.
Uncertainty Analysis.
ORNL is presently determining 1) uncertainties in instrument response, 2) uncertainties in parameters calculated from instru-ment response such spool piece mass flux and bundle power, 3) uncertainties in bundle steady state conditions from mass balances and energy balances ar.d
- 4) uncertainty in fluid conditions as predicted by computer codes. All four are a necessary part of data reduction if it is to be useful data for a broad spectrum of users.
It does not seem appropriate, however, for ORNL to tackle the uncertainties in code predictions.
Uncertainties in code predictions are difficult to estimate. This area is the goal of the entire code verification program. A great deal of effort ORNL and money is going into establishing the code uncertainties already.
should leave work in this area with Dr. S. Fabic and only have a minimal involve-ment to keep them knowledgeable.
1 I
a UNIVERSITY OF cal,lFORNI A-ilAtterhead for interdepartmental use)
~,...e,.
, -. _, - _.. _ - - - _